## wpiea2022200-print-pdf

## Source details

**Canonical URL:** [wpiea2022200-print-pdf](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022200-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022200-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022200-print-pdf.pdf.json)

---

### Overview and motivation
- SOEs support national interests, promote social objectives, and correct market failures (OECD 2015a).
- In the Middle East and Central Asia, SOE development is linked to government ownership of strategic assets (Ramirez Rigo and others 2021).
- SOEs concentrated in natural-monopoly sectors (gas, electricity, water, waste management, transportation) and present across the economy, including manufacturing.
- Two competing mechanisms for SOE performance:
  - Political connections confer advantages: subsidies, tariff protections, easier access to financing, lighter taxation, and stronger market power (Goldberg and others 2010; Chen and others 2017; Tao and others 2017; Cao and others 2019).
  - Political connections reduce incentives to operate efficiently: lower revenues, higher costs per employee, lower productivity (Dewenter and Malatesta 2001; European Commission 2016; Wang and Shailer 2018; Richmond and others 2019; Jurzik and Ruane 2021).
- Empirical evidence often finds SOEs have lower profitability due to below-cost pricing and employment objectives (Richmond and others 2019; IMF 2019).
- Prior studies: SOEs can avoid layoffs in downturns (Ki and Qotz 2019; Gu and others 2019); Ramirez Rigo and others 2021 find SOEs’ profitability suffered more during COVID-19 because they sheltered workers.

### Contribution of this paper
- Focus: Iran during international economic sanctions imposed in 2012.
- Data: Statistical Center of Iran (SCI) Survey of Manufacturing Firms (SMF) panel data to compare SOEs and private firms.
- Main empirical summary:
  - Sanctions reduced revenues, profits, and productivity of both SOEs and private firms in targeted industries.
  - Sanctions increased employment and wage levels only in SOEs.
  - SOEs’ limited workforce adjustment protected employment but resulted in sharper drops in profits and productivity compared to private firms, lowering dividends to the state.
  - SOEs introduce fiscal risks if the government continues to support employment.

### Distributional impacts and gender
- Prior work: sanctions reduced female labor force participation in Iran (Laudati and Pesaran 2019; Demir and Tabrizy 2019).
- This study: 2012 sanctions had a more significant impact on men overall.
- Male employment protected more than female employment in SOEs.
- Legal and institutional features:
  - Iranian Labor Law makes layoffs in SOEs extremely difficult.
  - Additional disincentives to hiring women: bans on women in "dangerous" jobs, need for husband’s consent before hiring a married woman, employer requirements for children’s care centers.

### Organization
- Paper organized with Section 2 data, Section 3 sanctions background and identification, Section 4 empirical strategy, Sections 5–6 estimates, Section 7 concluding.

---

### Data: sample, classification, and key statistics
- Sample: panel of Iranian manufacturing firms with 10 or more workers in 2009-2013 from SCI SMF.
- Coverage: between 12,556 and 14,168 firms each year; sample contains 130 4digit ISIC industries in 21 broad 2digit ISIC industries.
- Table coverage: over 12,801 firms annually between 2009-2013 in 130 4digit ISIC industries.
- SOE identification:
  - Firms labeled SOE when controlled and managed by the government (dataset allows identification of non-publicly listed controlled firms).
  - Identified 330 SOEs in the manufacturing sector during the sample period.
- Monetary variables in nominal Iranian rials (IRR); real values constructed by deflators:
  - Revenues deflated by industry-level PPI measures reported by the SCI.
  - Wages, intermediate inputs, investment, and capital stock deflated by PPI measures reported by the Central Bank of Iran (CBI).
  - Exports deflated by Export PPI measures produced by the CBI.

- SOE counts and shares:
  - 330 SOEs identified.
  - SOEs represent 4 percent of firms in manufacturing.
  - SOEs employ 11 percent of workers in the sector.
  - SOEs use 14 percent of capital.
  - SOEs contribute to more than 10 percent of the sector’s profits.

- Comparative means (values in billion tomans; 1 billion tomans = 10 billion IRR):
  - Revenues: POEs mean 76; SOEs mean 700.
  - Domestic revenues: POEs mean 72; SOEs mean 660.
  - Intermediate inputs: POEs mean 55; SOEs mean 490.
  - Capital: POEs mean 4.7; SOEs mean 40.
  - Labor (employees): POEs mean 77; SOEs mean 380.
  - Male Employee: POEs mean 68; SOEs mean 360.
  - Female Employee: POEs mean 9; SOEs mean 18.
  - College (or above) Employee: POEs mean 11; SOEs mean 61.
  - Non-College Employee: POEs mean 70; SOEs mean 410.
  - Male Employee - PL Unskilled: POEs mean 20; SOEs mean 66.
  - Male Employee - PL Skilled: POEs mean 20; SOEs mean 110.
  - Male Employee - NPL: POEs mean 11; SOEs mean 88.
  - Female Employee - PL Unskilled: POEs mean 3; SOEs mean 2.
  - Female Employee - PL Skilled: POEs mean 1; SOEs mean 2.
  - Female Employee - NPL: POEs mean 3; SOEs mean 9.
  - Wage (billion tomans): POEs mean 4.7; SOEs mean 38.

- Firm-size comparisons:
  - Revenues on average 9 times larger in SOEs compared to POEs.
  - Employment levels 5 times larger in SOEs compared to POEs.
  - Capital-to-labor ratio: SOEs 105 million IRR per worker; POEs 61 million IRR per worker.
  - Skill intensity: SOEs have 6.7 non-college graduates per college graduate; POEs have 6.3.

- Industry and gender patterns:
  - SOEs more present in transport and media; less in clothes and furniture.
  - In metals and vehicles industries, above 30 percent of revenues produced by SOEs; around one third of employees in metals, transport, media, and recycling employed in SOEs.
  - Gender employment ratios (2009-2013 average): in SOEs, 21 men per each female employee; in POEs, around 9 men per each female employee.
  - Wage gaps: men paid almost 10 times more than women in POEs on average; men paid 23 times more than women in SOEs on average.
  - Positions outside production line: in SOEs 10 men per every woman; in POEs 4 men per every woman.

---

### Sanctions background and identification
- Timeline and scope:
  - U.S. sanctions began in 1979; U.N. multilateral sanctions initiated in 2006.
  - EU sanctions culminated in a full embargo of Iranian oil and sanctions on financial institutions by 2012.
  - Sanctions lifted in 2015 under JCPOA; U.S. withdrew in 2018 and re-imposed sanctions thereafter.
  - Paper focuses on sanctions enacted between 2010 and 2013, peaking in 2012.

- Notable 2010–2012 policy actions:
  - 2010: CISADA intensified targeting petrochemical industry.
  - End-2011: Section 311 labeled entire financial system, including CBI, as a “jurisdiction of primary money laundering concern”.
  - End-2011: NDAA 2012 intensified measures.
  - 2012 US measures included:
    - Administration-ordered confiscation by domestic banks of all remaining Iranian assets.
    - Introduction of secondary sanctions against third countries buying Iranian oil and petroleum products.
    - August 2012: Congress passed TRA banning messaging services for financial transactions, banning shipping insurance provision, prohibiting repatriation of Iran’s oil receipts.
    - End-2012: IFCAP blacklisted entire energy, shipping, shipbuilding and port operating sectors; banned sale, supply and transfer of precious and semi-finished metals to Iran; prohibited provision of insurance and reinsurance to blacklisted entities; sanctioned state-owned radio and television agency.
    - 2012: DRIVE Act required automakers to certify no business with Iranian entities; resulted in Hyundai, Porsche, Peugeot and Fiat ending business with Iran.
  - 2012 EU measures included boycott of Iranian oil and petrochemical products effective 1 July 2012; ban on insurance for shipments; freeze of CBI assets; block exports of petrochemical equipment; March 2012 decision effectively cut Iranian banks off SWIFT.
  - Other countries joining US and EU sanctions included Japan, South Korea, Canada, Switzerland and Australia.

- Targeted industries and sample scope:
  - Analysis excludes financial and insurance sectors and the oil and petrochemical sector.
  - Industries identified as sanctioned (two-digit ISIC mapped to six four-digit ISIC in sample):
    - manufacture of motor vehicles;
    - manufacture of bodies (coachwork) for motor vehicles;
    - manufacture of trailers and semitrailers;
    - manufacture of parts and accessories for motor vehicles and their engines;
    - building and repairing of ships, building and repairing of pleasure and sporting boats;
    - manufacture of aircraft and spacecraft.
  - These industries account for around 20 percent of the manufacturing sector’s revenues over the sample period and 13 percent of manufacturing employment.
  - Targeted firms comprised about 25 percent of Iranian manufacturing firms’ revenue before the sanctions; half of these revenues were produced by SOEs.
  - The proportion of sanctioned SOEs in the analysis is a lower bound due to exclusions and reliance on SCI official ownership classification.

---

### Empirical strategy
- Difference-in-difference comparing averages 2009-2011 (pre-sanctions) to 2012-2013 (post-sanctions).
- Baseline regression (equation (1)):
  - lnYit = δSjt + αi + αt + εit
  - Yit: dependent variable (revenues, profits, firm-level productivity, employment levels, wages) for firm i at year t.
  - Sjt = 1 for firms in industry j sanctioned in or after 2012, 0 otherwise.
  - αi and αt are firm and year fixed effects.
  - Standard errors clustered at industry-year level.
- SOE heterogeneous-effects regression (equation (2)):
  - lnYit = βSjt + δSOEi × Sjt + SOEi + αi + αt + εit
  - SOEi = 1 if firm i is controlled and managed by the state before 2012.
  - β: impact of sanctions on targeted firms; β + δ: impact on targeted SOEs; δ: differential effect between POEs and SOEs.

---

### Aggregate impacts of sanctions (economy-wide manufacturing estimates)
- Table 2 estimates (dependent variables in log values; standard errors in parentheses):
  - Sanctionsjt on Profit: -0.358*** (0.062)
  - Sanctionsjt on Total Rev: -0.228*** (0.056)
  - Sanctionsjt on Emp: -0.157*** (0.015)
  - Sanctionsjt on Wage: -0.160*** (0.024)
- Observations and model fit:
  - Obs Profit: 54,324; Total Rev: 62,015; Emp: 65,062; Wage: 65,052
  - Adjusted R-squared Profit: 0.677; Total Rev: 0.820; Emp: 0.910; Wage: 0.872
- Reported percent translations (formula: (e^β − 1) * 100):
  - Sanctions reduced revenues by 20 percent and profits by 30 percent (as reported in text).
  - Employment fell by 14.5 percent and wages by 14.7 percent in targeted firms (as reported in text).

### Heterogeneous impacts: SOEs versus POEs (Table 2 lower panel)
- Coefficients (standard errors in parentheses):
  - Sanctionsjt: Profit -0.323*** (0.064); Total Rev -0.211*** (0.056); Emp -0.166*** (0.015); Wage -0.173*** (0.025)
  - SOEs × Sanctionsjt: Profit -0.524*** (0.139); Total Rev -0.279** (0.127); Emp 0.170*** (0.039); Wage 0.238*** (0.070)
- Observations: Profit 52,222; Total Rev 59,558; Emp 62,494; Wage 62,484
- Reported percent changes in text:
  - In SOEs: revenues decreased by almost 40 percent and profits were more than halved; employment levels and wages increased by 0.4 percent and 6.4 percent respectively.
  - In POEs: employment levels decreased by 15 percent and wages by 16 percent.
- Interpretation:
  - SOEs experienced larger falls in revenues and profits but maintained or increased employment and wages, consistent with labor rigidities and wage premia.
  - Iranian Labor Law provisions make dismissal difficult and expensive (citations to Sections 22 and 27).
  - Mehr-Afarin plan (Fall 2012) aimed to increase employees by half a million over four years, reinforcing employment maintenance incentives.

---

### Impacts on firm-level productivity (Table 3)
- Total sample impacts (dependent variables in log TFP-OP.ACF and TFP-LP.ACF; standard errors in parentheses):
  - Sanctionsjt: TFP-OP.ACF -0.240*** (0.073); TFP-LP.ACF -0.171** (0.072)
  - Observations: 50,384 for both; Adjusted R-squared 0.964 and 0.848 respectively.
- SOE heterogeneous impacts:
  - Sanctionsjt: -0.215*** (0.072) and -0.145** (0.071)
  - SOEs × Sanctionsjt: -0.372*** (0.108) and -0.387*** (0.110)
  - Observations: 49,326 for both; Adjusted R-squared 0.965 and 0.848.
- Reported percent changes in text:
  - Sanctions reduced firm-level productivity in targeted industries by 21 percent and 16 percent (two methods).
  - Productivity of SOEs dropped by around 40 percent compared to 13 percent for POEs after sanctions.
- Interpretation:
  - Larger productivity decline in SOEs linked to employment and wage rigidities and misallocation of labor (protected unskilled employment reducing productivity).

---

### Differential impacts across education, skill, and gender

- Education and skill (Table 4; dependent variables in log values; standard errors in parentheses):
  - Sanctionsjt effects:
    - College Emp: -0.071** (0.035)
    - Non-College Emp: -0.128*** (0.026)
    - Unskill Emp: -0.033 (0.041)
    - Skilled Emp: -0.220*** (0.025)
  - SOEs × Sanctionsjt:
    - College Emp: 0.086 (0.068)
    - Non-College Emp: 0.078 (0.059)
    - Unskill Emp: 0.288** (0.146)
    - Skilled Emp: -0.023 (0.075)
  - Observations:
    - College Emp: 25,631 / 24,732
    - Non-College Emp: 31,722 / 30,613
    - Unskill Emp: 59,773 / 57,382
    - Skilled Emp: 60,071 / 57,715
  - Reported percent changes in text:
    - Employment of college graduates dropped by 7 percent and non-college graduates by 12 percent after sanctions.
    - Sanctions reduced employment of skilled workers by 19.7 percent with no significant POE–SOE difference.
    - For unskilled workers, employment levels were 33.3 percent higher in SOEs than POEs after sanctions, indicating SOEs protected unskilled workers relatively more.
  - Interpretation:
    - Protection of unskilled workers in SOEs likely contributed to larger productivity declines relative to POEs.

- Gender differences (Table 5; dependent variables in log values; standard errors in parentheses):
  - Sanctionsjt effects:
    - Men Emp: -0.177*** (0.014)
    - Women Emp: -0.078*** (0.017)
    - Men Wages: -0.166*** (0.027)
    - Women Wages: -0.123*** (0.027)
  - SOEs × Sanctionsjt:
    - Men Emp: 0.169*** (0.043)
    - Women Emp: 0.065 (0.111)
    - Men Wages: 0.276*** (0.075)
    - Women Wages: 0.110 (0.197)
  - Observations:
    - Men Emp: 48,953; Women Emp: 44,250; Men Wages: 64,849; Women Wages: 43,821
    - Heterogeneous-panel observations slightly lower (e.g., Men Emp 47,053).
  - Reported percent changes in text:
    - Male employment decreased by 16.2 percent versus 7.5 percent for women.
    - Male wages dropped by 15.2 percent and female wages by 11.6 percent.
    - In SOEs after sanctions, male employment levels and wages were 18.4 percent and 31.8 percent higher than in POEs respectively.
    - Male employment levels dropped by 2 percent in SOEs when targeted; male wages showed an increase of 10 percent after the sanctions in SOEs.
  - Interpretation:
    - Men were hit harder overall by the sanctions, but in SOEs men fared better relative to POEs in both employment and wages.

---

### Mechanisms, robustness, and trade impacts
- Main channels of sanction impact:
  - Restriction of access to foreign intermediate inputs reduced profits, revenues, productivity, and employment.
  - Restriction of access to export markets decreased revenues for exporting firms.
- SOE-specific channels:
  - Labor market rigidities (Iranian Labor Law) and government employment support policies (Mehr-Afarin plan aiming to increase employees by half a million over four years) limited SOEs’ ability to adjust employment and wages downward.
  - Maintaining employment and paying wage premia in SOEs contributed to larger declines in revenues, profits and productivity vs POEs.
- Trade impacts (Appendix I; PPML estimation):
  - Sanctions reduced average firm-level exports and imports by 47 percent and 37 percent, respectively, compared to firms not subject to sanctions (Table A.1).
  - Number of importing firms dropped by 3 percent; impact on exporting firms was not significant.
  - Impacts on trade between POEs and SOEs not significantly different, except for a lower impact on exports for targeted SOEs.
  - Interpretation: SOE–POE differentials unlikely driven by changes in export/import channels; labor law constraints plausibly explain larger SOE productivity drops.
- Robustness (Appendix IV):
  - Including oil and petrochemical sector (Table A.2) yields sanctionsjt coefficients:
    - Profiti,t: -0.229**
    - Total Revi,t: -0.136**
    - Empi,t: -0.062*
    - Wagei,t: -0.081**
  - Heterogeneous SOE impacts in this specification:
    - SOEs × Sanctionsjt: -0.038, 0.012, 0.097**, 0.163***

- TFP estimation (Appendix II):
  - Method: Ackerberg et al. (2015) correction for OP/LP production function estimation; LP intermediate input function used.
  - Assumed Cobb-Douglas production function at 2-digit ISIC industry level; semi-parametric techniques and second-stage GMM; productivity modeled as Markov process; capital adjusts with a lag.

---

### Conclusion and policy implications
- Main findings:
  - Sanctions during 2010–2013, peaking in 2012, reduced revenues, profits and productivity in both SOEs and POEs in targeted manufacturing industries.
  - SOEs responded to revenue shocks by increasing employment and wages, reflecting limited labor adjustment and stronger employment protection; this insulated workers but amplified declines in SOEs’ profits and productivity.
  - Trade-offs: SOEs can stabilize employment during downturns but at the cost of fiscal risk and lower returns to the state.
  - Gendered outcomes: men were generally hit harder but were better supported in SOEs; wages increased for men and dropped for women; male employment contracted by less than women’s.
  - Labor law features and incentives to hire unskilled employees and men, and disincentives for hiring women, are associated with these outcomes.

- Policy recommendations:
  - Clarify SOEs' objectives if the government’s mandate is to protect workers during recessions; evaluate whether avoiding layoffs achieves objectives while minimizing fiscal costs.
  - Consider direct support to workers through unemployment insurance as an alternative to indirect support via SOE employment retention.
  - If indirect support via employment retention is used:
    - Treat workers in SOEs and private firms equally.
    - Set clear policies and conditions for when to provide support to firms to minimize fiscal risks.
  - To reduce discrimination against women and boost workforce inclusion:
    - Relax laws restricting women’s ability to work in all jobs.
    - Reduce obstacles to their employment (for example, the need to obtain husbands' consent for hiring married women).

*Source: IMF Working Paper — Introduction, Data, Empirical Strategy, Results, Conclusion (wpiea2022200-print-pdf).*

### Introduction ...........................................................................................................

### Introduction

### Overview and motivation
- SOEs are used to support national interests, promote social objectives, and correct market failures (OECD 2015a).
- In the Middle East and Central Asia, SOE development is linked to governments owning strategic assets, particularly in oil and gas, and to nationalization of major industries (Ramirez Rigo and others 2021).
- SOEs are concentrated in natural-monopoly sectors (gas, electricity, water, waste management, transportation) but are present across the economy, including manufacturing.
- Literature notes two competing mechanisms for SOE performance:
  - Political connections can confer advantages: subsidies, tariff protections, easier access to financing, lighter taxation, and stronger market power (Goldberg and others 2010; Chen and others 2017; Tao and others 2017; Cao and others 2019).
  - Political connections can reduce incentives to operate efficiently, producing lower revenues, higher costs per employee, and lower productivity (Dewenter and Malatesta 2001; European Commission 2016; Wang and Shailer 2018; Richmond and others 2019; Jurzik and Ruane 2021).
- Empirical evidence often finds SOEs have lower profitability due to below-cost pricing and employment objectives (Richmond and others 2019; IMF 2019).
- Prior studies show SOEs can offset downturns by avoiding layoffs (Ki and Qotz 2019; Gu and others 2019). Ramirez Rigo and others 2021 find SOEs’ profitability suffered more during COVID-19 because they sheltered workers.

### Contribution of this paper
- Focuses on Iran during international economic sanctions imposed in 2012.
- Uses Statistical Center of Iran (SCI) Survey of Manufacturing Firms (SMF) panel data to compare SOEs and private firms.
- Main empirical result summary:
  - Sanctions reduced revenues, profits, and productivity of both SOEs and private firms in targeted industries.
  - Sanctions increased employment and wage levels only in SOEs.
  - SOEs’ limited workforce adjustment protected employment but resulted in sharper drops in profits and productivity compared to private firms, lowering dividends to the state.
  - SOEs therefore introduce fiscal risks if the government continues to support employment.

### Distributional impacts and gender
- Prior work shows sanctions reduced female labor force participation in Iran (Laudati and Pesaran 2019; Demir and Tabrizy 2019).
- This study finds the 2012 sanctions had a more significant impact on men overall.
- Male employment was protected more than female employment in SOEs.
- Legal and institutional features explain heterogeneity:
  - Iranian Labor Law makes layoffs in SOEs extremely difficult.
  - Additional disincentives to hiring women: bans on women in "dangerous" jobs, need for husband’s consent before hiring a married woman, and requirements for employer-provided children’s care centers.

*This paper is organized with Section 2 presenting the data, Section 3 the sanctions background and identification of shocks, Section 4 the empirical strategy, Sections 5–6 estimates of sanctions’ impacts on economic and employment performance, and Section 7 concluding.*

---

### Data

### Sample and source
- Panel of Iranian manufacturing firms with 10 or more workers in the period 2009-2013.
- Data produced by the Statistical Center of Iran (SCI) in the Survey of Manufacturing Firms (SMF).
- The database covers between 12,556 and 14,168 firms each year.
- The sample contains 130 4digit ISIC level industries in 21 broad 2digit ISIC level industries.
- Table coverage: over 12,801 firms annually between 2009-2013 in 130 4digit ISIC industries.

### Firm classification and variables
- Firms labeled as SOE when controlled and managed by the government (dataset allows identification of non-publicly listed controlled firms).
- Identified 330 SOEs in the Iranian manufacturing sector during the sample period.
- Monetary variables available in nominal Iranian rials (IRR): profits, revenues, exports, intermediate inputs, wages, investment, and capital stock.
- Real values constructed by deflating:
  - Revenues by industry-level PPI measures reported by the SCI.
  - Wages, intermediate inputs, investment, and capital stock by PPI measures reported by the Central Bank of Iran (CBI).
  - Exports by Export PPI measures produced by the CBI.

### Key summary statistics and firm differences
- SOE counts and shares:
  - 330 SOEs identified.
  - SOEs represent 4 percent of firms in manufacturing.
  - SOEs employ 11 percent of workers in the sector.
  - SOEs use 14 percent of capital.
  - SOEs contribute to more than 10 percent of the sector’s profits.
- Comparative statistics (Table 1: means; values in billion tomans; 1 billion tomans = 10 billion IRR):
  - Revenues: POEs mean 76; SOEs mean 700.
  - Domestic revenues: POEs mean 72; SOEs mean 660.
  - Intermediate inputs: POEs mean 55; SOEs mean 490.
  - Capital: POEs mean 4.7; SOEs mean 40.
  - Labor (employees): POEs mean 77; SOEs mean 380.
  - Male Employee: POEs mean 68; SOEs mean 360.
  - Female Employee: POEs mean 9; SOEs mean 18.
  - College (or above) Employee: POEs mean 11; SOEs mean 61.
  - Non-College Employee: POEs mean 70; SOEs mean 410.
  - Male Employee - PL Unskilled: POEs mean 20; SOEs mean 66.
  - Male Employee - PL Skilled: POEs mean 20; SOEs mean 110.
  - Male Employee - NPL: POEs mean 11; SOEs mean 88.
  - Female Employee - PL Unskilled: POEs mean 3; SOEs mean 2.
  - Female Employee - PL Skilled: POEs mean 1; SOEs mean 2.
  - Female Employee - NPL: POEs mean 3; SOEs mean 9.
  - Wage (billion tomans): POEs mean 4.7; SOEs mean 38.
- Firm-size comparisons:
  - On average, revenues are 9 times larger in SOEs compared to POEs.
  - Employment levels are 5 times larger in SOEs compared to POEs.
  - Capital-to-labor ratio: SOEs 105 million IRR per worker; POEs 61 million IRR per worker.
  - Skill intensity: SOEs have 6.7 non-college graduates per college graduate; POEs have 6.3.

### Industry patterns and gender
- SOEs are more present in transport and media sectors, less in clothes and furniture.
- In metals and vehicles industries, above 30 percent of revenues are produced by SOEs; around one third of employees in metals, transport, media, and recycling are employed in SOEs.
- SOEs hire more employees than POEs in all manufacturing industries and employ more men relative to women.
- Gender employment ratios (2009-2013 average):
  - In SOEs, 21 men were hired per each female employee.
  - In POEs, around 9 men were hired per each female employee.
- Wage gaps:
  - Men were paid almost 10 times more than women in POEs on average.
  - Men were paid 23 times more than women in SOEs on average.
- Position-level gender disparities:
  - Positions outside the production line: in SOEs there were 10 men per every woman; in POEs there were 4 men per every woman.

---

### Background of Sanctions Against Iran and Identification of Sanctioned Industries

### Timeline and scope (2010–2013 focus)
- Sanctions on Iran span decades; U.S. sanctions began in 1979.
- U.N. multilateral sanctions initiated in 2006 after collapsed negotiations with France, Germany, and the United Kingdom.
- EU sanctions culminated in a full embargo of Iranian oil and sanctions on financial institutions and other sectors by 2012.
- Sanctions were lifted in 2015 under the JCPOA; the U.S. withdrew in 2018 and re-imposed sanctions thereafter.
- The paper focuses on sanctions enacted between 2010 and 2013, peaking in 2012.
- Sanctions in 2010–2013 were unprecedented in scope, targeting entire sectors and specific firms to weaken the economy and influence nuclear negotiations.

### Notable policy actions during the 2010–2013 episode
- 2010: U.S. intensified targeting of petrochemical industry and entities/individuals linked to the nuclear program (Comprehensive Iran Sanctions, Accountability and Divestment Act (CISADA)).
- Late 2011: Following an IAEA report, sanctions intensified further.
- Under Section 311 of the USA Patriot Act, the U.S. labeled the entire financial system, including the Central Bank of Iran (CBI), as a “jurisdiction of primary money laundering concern”.
- End of 2011: Obama signed the National Defense Authorization Act (NDAA 2012), intensifying measures.

---

### Empirical Strategy and Key Impacts (overview)

### Identification strategy
- Uses firm-level panel data for 2009–2013 across 130 4digit ISIC industries.
- Sanction shocks identified based on industry targeting and timing in 2010–2013, with peak effects in 2012.

### Main estimated impacts (summary from empirical sections)
- Sanctions’ effects on firm outcomes (reported in subsequent empirical sections and tables):
  - Reduced revenues and profits for both SOEs and POEs in targeted industries.
  - Reduced firm-level productivity (TFP) in targeted industries.
  - Increased employment and wages in SOEs but not in POEs.
  - Larger drops in profits and productivity in SOEs relative to POEs due to limited workforce adjustment.
- Implications:
  - SOEs’ employment protection functions as a labor market stabilizer but imposes fiscal and productivity costs.
  - Government faces fiscal risks when SOEs are used to shield employment during downturns.

---

### Impacts of Sanctions on Different Types of Employees in SOEs and POEs

### Heterogeneous impacts by gender and skill
- Overall sanctions impact:
  - Sanctions had a more significant impact on men overall.
  - Male employment was protected more than female employment in SOEs.
- Institutional drivers:
  - Iranian Labor Law makes layoffs in SOEs difficult.
  - Employment hurdles for women include bans from “dangerous” jobs (Section 75), half-hour nursing breaks treated as work hours (Section 78), and Article 1117 allowing a husband to prevent his wife’s employment if it conflicts with his dignity or the family’s interests.
- Skill and position heterogeneity (referenced in tables):
  - Differences across college versus non-college employees and across production-line (PL) and non-production-line (NPL) positions are analyzed in the paper (see Table 4 and Table 5 for detailed estimates).

---

### Conclusion

- Sanctions during 2010–2013, peaking in 2012, reduced revenues, profits, and productivity for both SOEs and POEs in targeted manufacturing industries in Iran.
- SOEs responded to revenue shocks by increasing employment and wages, reflecting limited labor adjustment and stronger employment protection; this insulated workers but amplified declines in SOEs’ profits and productivity.
- The trade-off implies SOEs can stabilize employment during downturns but at the cost of fiscal risk and lower returns to the state.
- Gendered labor market outcomes show male employment was relatively more protected, and SOEs exhibited larger male–female disparities in employment and wages, shaped by legal constraints on female labor participation.

*Source: IMF Working Paper — Introduction (wpiea2022200-print-pdf).*

### section 1425 of which barred foreign banks to process oil receipts through the CBI. Furthermore, the

### wpiea2022200-print-pdf - section 1425 of which barred foreign banks to process oil receipts through the CBI. Furthermore, the

### Sanctions timeline and scope (2012)
- US measures in 2012 included:
  - Administration-ordered confiscation by domestic banks of all remaining Iranian assets.
  - Introduction of secondary sanctions against third countries buying Iranian oil and petroleum products.
  - August 2012: Congress passed the Iran Threat Reduction and Syria Human Rights Act (TRA) that:
    - banned provision of messaging services for conducting financial transactions;
    - banned shipping insurance provision;
    - prohibited the repatriation of Iran’s oil receipts.
  - End-2012: Iran Freedom and Counter-Proliferation Act (IFCAP):
    - blacklisted the entire energy, shipping, shipbuilding and port operating sectors;
    - banned sale, supply and transfer of precious and semi-finished metals to Iran;
    - prohibited provision of insurance and reinsurance to blacklisted entities;
    - sanctioned the state-owned radio and television agency.
  - 2012: DRIVE Act (proposed by UANI) required automakers to certify no business with Iranian entities to be eligible for US government contracts or financial assistance; resulted in Hyundai, Porsche, Peugeot and Fiat ending business with Iran and stopping shipments of auto parts.
- EU measures in 2012 included:
  - Boycott of Iranian oil and petrochemical products effective as of 1 July 2012;
  - Ban on insurance for their shipments; freeze of CBI assets; block exports of petrochemical equipment and technology; bar trade of diamonds and precious metals with Iran (Council Conclusions on Iran, 23 January 2012).
  - March 2012 decision preventing Iran’s access to financial messaging services for clearing banking transactions (effectively cutting off Iranian banks from the SWIFT network).
- Other countries joining US and EU sanctions included Japan, South Korea, Canada, Switzerland and Australia.

### Targeted industries and sample scope
- Analysis excludes:
  - financial and insurance sectors (database covers only manufacturing firms);
  - oil and petrochemical sector (excluded to avoid obfuscation from oil price dynamics).
- Industries identified as sanctioned (two-digit ISIC level) and mapped to six four-digit ISIC industries in sample:
  - manufacture of motor vehicles;
  - manufacture of bodies (coachwork) for motor vehicles;
  - manufacture of trailers and semitrailers;
  - manufacture of parts and accessories for motor vehicles and their engines;
  - building and repairing of ships, building and repairing of pleasure and sporting boats;
  - manufacture of aircraft and spacecraft.
- These industries account for:
  - around 20 percent of the manufacturing sector’s revenues over the sample period;
  - 13 percent of manufacturing employment over the sample period.
- Targeted firms comprised:
  - about 25 percent of Iranian manufacturing firms’ revenue before the sanctions;
  - half of these revenues were produced by SOEs.
- The proportion of sanctioned SOEs in the analysis is interpreted as a lower bound due to exclusions (petrochemical/oil and nonmanufacturing industries) and reliance on SCI official ownership classification.

### Empirical strategy
- Difference-in-difference specification comparing averages 2009-2011 (pre-sanctions) to 2012-2013 (post-sanctions).
- Baseline regression (equation (1)):
  - lnYit = δSjt + αi + αt + εit
  - Yit: dependent variable (revenues, profits, firm-level productivity, employment levels, wages) for firm i at year t.
  - Sjt = 1 for firms in industry j sanctioned in or after 2012, 0 otherwise.
  - αi and αt are firm and year fixed effects.
  - Standard errors clustered at industry-year level.
- SOE heterogeneous-effects regression (equation (2)):
  - lnYit = βSjt + δSOEi × Sjt + SOEi + αi + αt + εit
  - SOEi = 1 if firm i is controlled and managed by the state before 2012.
  - β: impact of sanctions on targeted firms; β + δ: impact on targeted SOEs; δ: differential effect between POEs and SOEs.

### Aggregate impacts of sanctions on manufacturing firms (Table 2)
- Estimated coefficients (dependent variables are in log values):
  - Sanctionsjt on Profit: -0.358*** (standard error 0.062)
  - Sanctionsjt on Total Rev: -0.228*** (0.056)
  - Sanctionsjt on Emp: -0.157*** (0.015)
  - Sanctionsjt on Wage: -0.160*** (0.024)
- Observations and model fit:
  - Obs Profit: 54,324; Total Rev: 62,015; Emp: 65,062; Wage: 65,052
  - Adjusted R-squared Profit: 0.677; Total Rev: 0.820; Emp: 0.910; Wage: 0.872
- Translation to percent impacts (formula provided in source: (e^β − 1) * 100):
  - Sanctions reduced revenues by 20 percent and profits by 30 percent (as reported in text).
  - Employment fell by 14.5 percent and wages by 14.7 percent in targeted firms (as reported in text).

### Heterogeneous impacts: SOEs versus POEs (Table 2 lower panel)
- Coefficients:
  - Sanctionsjt: Profit -0.323*** (0.064); Total Rev -0.211*** (0.056); Emp -0.166*** (0.015); Wage -0.173*** (0.025)
  - SOEs × Sanctionsjt: Profit -0.524*** (0.139); Total Rev -0.279** (0.127); Emp 0.170*** (0.039); Wage 0.238*** (0.070)
- Observations: Profit 52,222; Total Rev 59,558; Emp 62,494; Wage 62,484
- Reported percent changes in text:
  - In SOEs: revenues decreased by almost 40 percent and profits were more than halved; employment levels and wages increased by 0.4 percent and 6.4 percent respectively.
  - In POEs: employment levels decreased by 15 percent and wages by 16 percent.
- Interpretation:
  - SOEs experienced larger falls in revenues and profits but maintained or increased employment and wages, consistent with literature on SOE labor rigidities and wage premia.
  - Iranian Labor Law provisions make dismissal difficult and expensive, contributing to SOE rigidity (citations to Sections 22 and 27 of the law).
  - Mehr-Afarin plan (Fall 2012) aimed to support employment with objective of increasing employees by half a million over four years, which likely reinforced incentives to maintain employment.

### Impacts on firm-level productivity (Table 3)
- Total sample impacts (dependent variables in log TFP-OP.ACF and TFP-LP.ACF):
  - Sanctionsjt: TFP-OP.ACF -0.240*** (0.073); TFP-LP.ACF -0.171** (0.072)
  - Observations: 50,384 for both; Adjusted R-squared 0.964 and 0.848 respectively.
- SOE heterogeneous impacts:
  - Sanctionsjt: -0.215*** (0.072) and -0.145** (0.071)
  - SOEs × Sanctionsjt: -0.372*** (0.108) and -0.387*** (0.110)
  - Observations: 49,326 for both; Adjusted R-squared 0.965 and 0.848.
- Reported percent changes in text:
  - Sanctions reduced firm-level productivity in targeted industries by 21 percent and 16 percent (two methods).
  - Productivity of SOEs dropped by around 40 percent compared to 13 percent for POEs after sanctions.
- Interpretation:
  - Larger productivity decline in SOEs linked to employment and wage rigidities and potential misallocation of labor (protected unskilled employment contributing to lower productivity).

### Differential impacts across employee types (education, skill, gender)
- Education and skill (Table 4; dependent variables in log values):
  - Sanctionsjt effects:
    - College Emp: -0.071** (0.035)
    - Non-College Emp: -0.128*** (0.026)
    - Unskill Emp: -0.033 (0.041)
    - Skilled Emp: -0.220*** (0.025)
  - SOEs × Sanctionsjt:
    - College Emp: 0.086 (0.068)
    - Non-College Emp: 0.078 (0.059)
    - Unskill Emp: 0.288** (0.146)
    - Skilled Emp: -0.023 (0.075)
  - Observations:
    - College Emp: 25,631 (full sample) / 24,732 (heterogeneous panel)
    - Non-College Emp: 31,722 / 30,613
    - Unskill Emp: 59,773 / 57,382
    - Skilled Emp: 60,071 / 57,715
  - Reported percent changes in text:
    - Employment of college graduates dropped by 7 percent and non-college graduates by 12 percent after sanctions.
    - Sanctions reduced employment of skilled workers by 19.7 percent with no significant POE–SOE difference.
    - For unskilled workers, employment levels were 33.3 percent higher in SOEs than POEs after sanctions, indicating SOEs protected unskilled workers relatively more.
  - Interpretation:
    - Protection of unskilled workers in SOEs likely contributed to SOEs’ larger productivity declines relative to POEs.
- Gender differences (Table 5; dependent variables in log values):
  - Sanctionsjt effects:
    - Men Emp: -0.177*** (0.014)
    - Women Emp: -0.078*** (0.017)
    - Men Wages: -0.166*** (0.027)
    - Women Wages: -0.123*** (0.027)
  - SOEs × Sanctionsjt:
    - Men Emp: 0.169*** (0.043)
    - Women Emp: 0.065 (0.111)
    - Men Wages: 0.276*** (0.075)
    - Women Wages: 0.110 (0.197)
  - Observations:
    - Men Emp: 48,953; Women Emp: 44,250; Men Wages: 64,849; Women Wages: 43,821
    - Heterogeneous-panel observations slightly lower (e.g., Men Emp 47,053).
  - Reported percent changes in text:
    - Male employment decreased by 16.2 percent versus 7.5 percent for women.
    - Male wages dropped by 15.2 percent and female wages by 11.6 percent.
    - In SOEs after sanctions, male employment levels and wages were 18.4 percent and 31.8 percent higher than in POEs respectively.
    - Male employment levels dropped by 2 percent in SOEs when targeted; male wages showed an increase of 10 percent after the sanctions in SOEs.
  - Interpretation:
    - Men were hit harder overall by the sanctions, but in SOEs men fared better relative to POEs in both employment and wages.

### Mechanisms and interpretation
- Main channels of sanction impact:
  - Restriction of access to foreign intermediate inputs reduced profits, revenues, productivity, and employment in targeted industries.
  - Restriction of access to export markets decreased revenues in foreign markets and production for exporting firms.
- SOE-specific mechanisms:
  - Labor market rigidities (Iranian Labor Law) and government employment support policies (e.g., Mehr-Afarin plan aiming to increase employees by half a million over four years) limited SOEs’ ability to adjust employment and wages downward.
  - Maintaining employment and paying wage premia in SOEs contributed to larger declines in revenues, profits and productivity in SOEs relative to POEs.
- Robustness and additional tests:
  - Table (A.2) in Appendix IV shows robustness including oil and petrochemical firms (mentioned but not reproduced here).
  - Appendix I tests whether different trade patterns explain SOE–POE differential impacts and concludes sanctions did not have significantly different effects on SOEs’ exports and imports compared to POEs.

_Italic: Source — wpiea2022200-print-pdf (excerpts provided)_

### Conclusion

### Conclusion

### Main findings
- Sanctions depressed revenue, profits and productivity in both SOEs and private firms.
- Employment levels and wages increased in SOEs, suggesting they were used to protect employees during periods of economic crisis; this had adverse implications for Iran’s fiscal position.
- Low-skilled males were the most protected after the sanctions, which could help explain the productivity gap between SOEs and private firms.
- Gender dynamics:
  - Men were generally hit harder by the sanctions, but were better supported in SOEs compared to women.
  - Wages increased for men and dropped for women.
  - Male employment contracted by less than women’s after the shock.
- Labor law features supporting employment in SOEs, introducing incentives in hiring unskilled employees and men, and disincentives in hiring women, are associated with these outcomes.

### Trade impact (Appendix I)
- Estimation approach: intensive margins estimated using Poisson Pseudo Maximum Likelihood (PPML) due to many zero trade values.
- Table A.1 results (selected):
  - Sanctions reduced the average firm-level exports and imports by 47 percent and 37 percent, respectively, compared to firms not subject to sanctions.
  - The number of importing firms dropped by 3 percent, while the impact on exporting firms was not significant.
  - Impacts on trade between POEs and SOEs were not significantly different, except for a lower impact on exports for targeted SOEs.
- Interpretation: differences between SOEs and POEs in the aftermath of sanctions are unlikely to be driven by changes in firms’ export and import channels; labor law constraints on SOEs (difficulty laying off employees or decreasing real wages, incentives to hire during downturns) are likely contributors to the larger productivity drop in SOEs.

### Robustness checks (Appendix IV)
- Table A.2: Impact of Sanctions on Firms’ Profits, Revenues, Employment Levels, and Wages (Including the Oil and Petrochemical Sector) — dependent variables in log:
  - Sanctionsjt coefficients:
    - Profiti,t: -0.229**
    - Total Revi,t: -0.136**
    - Empi,t: -0.062*
    - Wagei,t: -0.081**
  - Heterogeneous impacts on SOEs (Sanctionsjt and SOEs × Sanctionsjt coefficients):
    - Sanctionsjt: -0.223***, -0.136**, -0.070*, -0.095**
    - SOEs × Sanctionsjt: -0.038, 0.012, 0.097**, 0.163***

### TFP estimation (Appendix II)
- Methodology: follow Ackerberg et al. (2015) to correct for functional dependence problems in OP/LP production function estimation; LP intermediate input function used where firm’s raw material inputs proxy unobserved productivity shocks.
- Functional form: Cobb-Douglas production function assumed for each industry; parameters estimated at 2-digit ISIC industry level.
- Estimation details: semi-parametric techniques for variable inputs; second-stage GMM to estimate input coefficients under identifying assumptions; productivity modeled as a Markov process and capital adjusts with a lag.

### Policy implications and recommendations
- SOEs' inability to flexibly adjust workforce during downturns harmed profits and productivity and could create the need for government intervention and recapitalization.
- If the government’s mandate is to protect workers during recessions, clarify SOEs' objectives and evaluate whether avoiding layoffs achieves those objectives while minimizing fiscal costs.
- Consider direct support to workers through unemployment insurance (as proposed in Richmond and others 2019) as an alternative to indirect support via employment retention in SOEs.
- If indirect support via employment retention is chosen:
  - Treat workers in SOEs and private firms equally.
  - Set clear policies and conditions for when to provide support to firms to minimize fiscal risks.
- To reduce discrimination against women and boost their workforce inclusion:
  - Relax the law to allow women to work in all jobs.
  - Reduce obstacles to their employment (for example, the need to obtain their husbands' consent to be able to hire them).

*Source: Conclusion and Appendices of "How Do State-Owned Enterprises Adjust During Downturns? Evidence from Iranian Manufacturing Firms", IMF Working Paper No. WP/2022/200*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022200-print-pdf.pdf_